PJM’s first large load process produced 12 proposals and no consensus. Now an accelerated procurement is addressing a 6.8 GW capacity shortfall.
PJM’s latest capacity auction left the RTO 6.8 GW below its Reliability Requirement for the 2028/29 delivery year, increasing pressure on the grid operator as it develops rules for rapidly growing large-load demand, including data centers, and seeks additional power supply.
PJM has responded with a proposed Reliability Backstop Procurement, an accelerated process intended to secure additional generation and address the shortfall. PJM filed the proposal with the Federal Energy Regulatory Commission July 31 after its board advanced the plan.
For data center operators, PJM’s large-load decisions increasingly determine whether proposed projects can secure the power they need to move forward, how much generation they may need to bring or contract for, and how much certainty they can have around long-term electricity costs. The approach is also testing a lesson from PJM’s first major attempt to address large load growth through its stakeholder process.
That process, launched in 2025, produced 12 proposals but no consensus. With none receiving the required stakeholder support, PJM’s board used the record to set its own policy direction and direct further work.
The question now is whether PJM can move faster while keeping policy direction, stakeholder implementation and the evidence supporting major planning assumptions clearly separated.
Twelve Proposals, No Decision #
A recently released transcript of a July 23 FERC technical conference on PJM governance shows how sharply participants disagreed over the stakeholder process.
PJM’s first Critical Issue Fast Path process for large loads produced 12 proposals, but all 12 were voted down by stakeholders and none advanced as a recommendation to the PJM Board of Managers.
Pennsylvania Deputy Secretary of Policy Jacob Finkel, speaking at the conference, recalled the reaction when he returned to PJM’s Valley Forge headquarters for a postmortem.
“The members went around patting themselves on the back for what a thoughtful, deliberative, creative process they had just gone through,” Finkel said.
He saw it differently.
“This is a failure,” Finkel said. “You have not advanced the ball.”
The problem was not a lack of ideas, said Neil Osnato, founder of Persistence Analytics Group.
The stakeholder process was being asked to resolve a policy question involving participants with materially different economic interests, downside exposures and views of who should pay, Osnato said.
“At some point, more alternatives do not create consensus,” Osnato said. “They create more ways to avoid making the underlying decision.”
Joseph Bowring, president of Monitoring Analytics and PJM’s independent market monitor, said the stakeholder process itself remains useful, although it could be made more efficient.
“Stakeholders were generally pursuing their narrow self interest and ignoring the broader public interest in efficient, competitive markets that assign costs and risks to those that cause them, in this case the data center load,” Bowring told Data Center Knowledge.
The proposals addressed different pieces of the large load problem, including load forecasting, new generation, interconnection, reliability and the treatment of customers that could not be served under PJM’s traditional resource adequacy framework.
PJM’s board ultimately used the CIFP record to develop its large-load framework, incorporating ideas from multiple stakeholder proposals. The January decision addressed load forecasting, bring your own new generation arrangements, expedited generation interconnection and a reliability backstop.
The Board Sets Direction #
Osnato said the outcome does not mean the stakeholder process had no value.
It generated information, exposed competing positions and gave the board a menu of possible approaches, he said.
But the process did not make the final policy decision.
“The stakeholder process functioned as an input mechanism, but failed as the final decision mechanism on this particular issue,” Osnato said.
Abe Silverman, assistant research scholar at Johns Hopkins University’s Ralph O’Connor Sustainable Energy Institute, made a similar distinction at the FERC conference.
The problem was not necessarily the number of committees in PJM’s stakeholder process, Silverman said.
“It’s when we ask committees to decide policy,” he said.
Large load integration should start with PJM, its board and the states establishing the policy direction, Silverman said. Stakeholders could then work through the technical and implementation questions.
“We would have saved so many hours of meetings,” he said.
Pamela Quinlan, executive director of the New Jersey Energy Resilience Bank, made a similar point while cautioning against abandoning PJM’s technical stakeholder process.
Direction and leadership should come from PJM leadership and the board, Quinlan said, while stakeholders provide the technical expertise needed to work through implementation.
PJM’s July Base Residual Auction procured 138,318 MW through the Reliability Pricing Model. Including fixed-resource-requirement capacity, PJM reported total RTO procured capacity of 149,181.6 MW UCAP, which was 6,831.3 MW UCAP below its Reliability Requirement for the 2028/29 delivery year.
PJM’s RBP proposal is designed to address that shortfall. The procurement is targeted to begin Sept. 30, subject to FERC approval.
The RBP would secure resources for terms of up to 15 years, with a maximum weighted-average willingness to pay of $555 per MW-day. Qualifying new resources covered by bilateral agreements or otherwise brought to the system would reduce the amount PJM procures through the backstop.
Customers Need Power Path #
The reliability gap has direct consequences for the companies driving much of PJM’s new demand.
Google’s Brian George, senior lead for U.S. energy markets, told FERC that his company has billions of dollars invested in PJM and operates data centers throughout the region.
Google has participated in the large load process despite not being a PJM member, working with members and other nonmembers on proposals intended to address the resource adequacy problem.
But George said the company needs greater certainty before committing capital to additional power resources.
The company is expected to make investments in new generation as part of the federal ratepayer protection effort.
“We can’t do that without the certainty that the market rules are going to exist in such a way that we know that investment is going to be worthwhile,” George said.
The RBP is intended to create a mechanism for bringing additional generation into the system while PJM works through its broader large-load framework.
PJM proposed a central procurement to address the 6.8 GW capacity shortfall from the 2028/29 Base Residual Auction. The procurement would run in September and October, with results expected in December, and would give new generation resources a fixed price for 15 years. PJM also proposed reducing the procurement target for large loads already matched with bilateral supply, approved integrated resource plan supply or demand-side participation.
Bowring said the RBP would shift the cost of serving data center load to other customers.
“The RBP process would require other customers to pay for data center load, despite PJM’s assertions to the contrary,” Bowring said. “The RBP approach is going in the wrong direction.”
Faster Decisions, Stronger Evidence #
The 6,831 MW auction shortfall is now the basis for a procurement that could involve substantial new generation investment and long-lived infrastructure commitments.
Osnato said that makes the evidentiary question even more important.
Before the shortfall figure drives procurement, cost allocation or infrastructure decisions, PJM should be able to identify the assumptions behind the number, the evidence supporting those assumptions and which large-load additions are executable and durable, Osnato said.
PJM should also establish which dependencies must remain true and what changes would require the calculation to be revisited, he said.
The goal should not be to slow PJM down.
“Speed matters,” Osnato said. “But speed and verification should not be treated as opposites.”
The same principle applies to PJM’s broader large-load planning.
“Before a large load representation becomes an infrastructure planning assumption,” Osnato said, “someone should be able to answer: Who proved it?”
The question is increasingly consequential as PJM tries to accommodate a wave of new data center demand while addressing the capacity shortfall.
PJM expects roughly 70 GW of new large load could come to the system by 2038.
At the same time, PJM is turning to an accelerated procurement process to address the immediate reliability shortfall.
Authority Or Execution? #
The disagreement over how to fix PJM’s process extends to whether the grid operator needs more authority.
Mason Emnett, senior vice president for public policy at Constellation, said the problems drawing the most attention, including resource adequacy and infrastructure, are already areas where PJM has Section 205 filing authority.
Changing that authority would not necessarily address the problem.
“PJM is the chef,” Emnett said. “It has different tools in its toolbox. And we don’t believe that those tools need to be changed; we believe they should be used.”
Emnett said PJM had already spent substantial time working through large load solutions and should not simply begin another lengthy process.
“That is about leadership and execution of responsibilities,” he said.
Osnato said the RBP does not fundamentally change his view of the first CIFP.
If anything, he said, the new procurement sharpens the lesson. The RBP addresses a more immediate problem, with PJM facing a stated capacity shortfall and an accelerated process intended to produce a backstop solution. But the governance risk remains if stakeholders are again being asked to determine core policy choices rather than implement a direction established by PJM and its board.
“The lesson from the first CIFP is not that PJM needs fewer ideas. It is that policy direction has to come before stakeholder implementation, and faster decisions need stronger evidence, not weaker evidence,” Osnato said.
The first large-load CIFP showed that a stakeholder process could generate a dozen serious proposals without producing a decision.
The test for PJM now is whether it can move faster without lowering the evidentiary standard for the assumptions behind those decisions.